Receptor conversion in distant breast cancer metastases. Breast cancer metastases: A spitting image of their primary?

Transcription

1 Receptor conversion in distant breast cancer metastases Breast cancer metastases: A spitting image of their primary?

2 Introduction Breast cancer is the leading cause of female cancer death worldwide (13,000 new cases yearly in The Netherlands) One third of patients will develop distant metastases (bone, liver, lungs, skin, brain) Distant metastases necessitate systemic therapy Chemotherapy Hormonal therapy targeting estrogen receptor (ERα) progesterone receptor (PR) Antibody therapies targeting e.g. human epidermal growth factor receptor 2 (HER2)

3 Systemic therapy for metastases Therapy choice now personalized on the basis of immunophenotype of primary tumor Metastases generally not biopsied (inconvenient/inaccessible/deemed unnecessary) Previous studies* indicate that immunophenotype of breast cancer metastases may differ from the primary tumor ( receptor conversion ) -ER % conversion rates: 12-57% -PR % conversion rates: % -HER2 % conversion rates: 0-44% This means that many metastatic patients may not get the appropriate systemic treatment when based on the primary tumor! *Kuukasjärvi T et al. J Clin Onc 1996 Tanner M et al. Cancer Res 2001 Broom RJ et al. Anticancer res 2009 Liedtke et al. Ann Oncol 2009

4 Aim So: Studies so far are small Just one metastatic site Ligand-binding assay used TMA Bone metastases included (false negativity?) Original primary tumor staining used! Aim: To study frequency of receptor conversion for ER, PR, and HER2: Large group of breast cancer metastases Different sites (brain, liver, lung, skin, GI) No bone metastases New stainings, one protocol Full sections

5 Material and Methods 236 paired samples of primary breast cancer and corresponding distant metastases ( Dutch breast cancer metastases consortium ) Sites: 44 brain, 43 lung, 63 liver, 79 skin, 7 gastro-intestinal Immunohistochemistry (IHC) for ERα, PR and HER2 HER2 silver in situ hybridization (SISH) was done in cases of IHC conversion or when primary tumors or metastases were IHC 2+ Whole sections, fresh immunohistochemical staining

6 Results

7 Results a primary c metastasis ER liver ER c d brain PR e f skin HER2 g h liver

8 Results: conversion for ER and PR 10% threshold Metastasis Primary tumor - + Total ERα - 79 (33.9%) 7 (3.0%) (7.2%) 130 (55.8%) 147 PR - 92 (39.5%) 12 (5.1%) (24.9%) 71 (30.5%) 129 In total: conversion rate for ER 10.2%, for PR 30%

9 Results: conversion for ER and PR 1% threshold Metastasis Primary tumor - + Total ERα - 47 (20.2%) 12 (5.2%) (9.9%) 151 (64.8%) 174 PR - 33 (14.2%) 27 (11.6%) (21.0%) 124 (52.2%) 173 In total: conversion rate for ER 15.1%, for PR 32.6%

10 Results: HER2 ERα, PR and HER2 expression by IHC in paired primary tumors and distant breast cancer metastases Metastasis Primary tumor - + Total HER2 IHC (77.2%) 6 (2.6%) (2.6%) 41 (17.6%) 47 Of the 12 cases with HER2 conversion by IHC, 5 also showed conversion by SISH. One further case showed conversion by SISH but not by IHC.

11 Results: conversion rates per metastatic site % conversion ERα PR HER2 N N (%) 95% CI N (%) 95% CI N(%) 95% CI Brain 44 6 (13.7) (36.3) (2.3) Lung 43 4 (9.4) (18.6) (4.7) Liver 63 8 (12.7) (41.2) (9.5) Skin 79 5 (6.3) (21.5) (2.5) Gastrointestinal 7 1 (14.3) (42.9) (14.3) 0-41 Receptor conversion seemed to occur mostly in liver and brain metastases for ER and PR, and in liver metastases for HER2 (GI?)

12 Results: ER and PR Clinically, tamoxifen is given when primary tumor is either ER+ or PR+, and withheld when ER and PR are both negative So, when basing tamoxifen therapy on immunophenotype of primary tumor, clinically there are 2 problematic situations: ER or PR+ ER-/PR- OR ER-/PR- ER+ or PR+ 10% threshold: ER- primary and ER+ or PR+ metastases: 3.4% ER+ or PR+ primary and ER-/PR- metastases: 10.7% 14% 1% threshold: ER-/PR- primary and ER+ or PR+ metastases: 8.2% ER+ or PR+ primary and ER-/PR- metastases: 12.4% 21%

13 Results: Heterogeneity between metastases 40 cases with multiple metastases, 26 analyzed Receptor homogeneous heterogeneous ER 23 3 PR 19 7 HER2 26 0

14 Results: receptor conversion due to tamoxifen treatment? % conversion Receptor Threshold Hormonal therapy No yes ER 10% no yes % no yes 95 5 PR 10% no yes % no yes 80 20

15 Results: Prognostic value of ER conversion 1% threshold 1,0 0,8 Cum survival 0,6 0,4 p< ,2 no conversion 0,0 conversion - --> + conversion + --> Survival time (days)

16 Results: Prognostic value of PR conversion 1% threshold 1,0 0,8 p=0.011 Cum survival 0,6 0,4 0,2 no conversion 0,0 conversion + --> - conversion - --> Survival time (days)

17 Results: Prognostic value of ER/PR conversion 1% threshold 1,0 0,8 p<0.001 Cum survival 0,6 0,4 no conversion 0,2 ER or PR conversion 0,0 ER and PR conversion Survival time (days)

18 Results: Prognostic value of HER2 conversion IHC ISH 1,0 1,0 0,8 p= ,8 p=0.374 Cum survival 0,6 0,4 no conversion Cum Survival 0,6 0,4 no conversion conversion conversion 0,2 0,2 0,0 0, Survival time (days) Follow up in days

19 Conclusions: not a spitting image! Frequency of receptor conversion in distant breast cancer metastases: ERα10-15%, PR 30-33%, HER2 3-5% Receptor conversion is seen especially in brain and liver metastases In approximately 16-26% of patients such conversion would have direct consequences for the therapeutic regimen Adjuvant tamoxifen treatment does not seem to induce receptor conversion Heterogeneity between metastases from the same patient seems limited

20 Conclusions Receptor conversion has negative prognostic impact These results underline the importance of having distant breast cancer metastases biopsied when possible Alternatively, molecular imaging strategies should be explored for sites that are difficult to biopsy PET imaging with 89 Zr-Trastuzumab PET imaging with 18 Fluoroestradiol Courtesy dr. Elisabeth de Vries

21 Acknowledgements UMC Utrecht: Laurien Hoefnagel Petra van der Groep Elsken van der Wall AMC Amsterdam Marc van de Vijver UMC St Radboud/CWZ Pieter Wesseling Iris Nagtegaal NKI-AVL Jelle Wesseling Lab for pathology Dordrecht: Pieter Westenend UMCG: Joost Bart VUmc: Paul van der Valk St Antonius Hospital: Cees Seldenrijk Diakonessenhuis Utrecht: Joost Oudejans Medisch Centrum Alkmaar Henk Jan van Slooten